The Quarterly Outlook as Consensus Laundering

In November 2025, J.P. Morgan Private Bank released its 2026 outlook and announced three themes defining “a new era”: artificial intelligence, global fragmentation, inflation. Six months earlier, the same desk had titled its mid-year outlook Comfortably Uncomfortable. Three years before that, none of those themes ranked as strategic priorities in the major banks’ documents. What changed was not reality; what changed was what had become socially acceptable to say.

Bloomberg’s compilation of 2026 outlooks aggregates more than sixty institutions and over seven hundred predictions, and the optimism on AI is described as “almost universal”. Fidelity calls it “the defining theme for equity markets”. BlackRock speaks of a mega force that will continue to beat tariffs and traditional macro drivers. JPMorgan Wealth Management states that “the biggest risk, in our view, is not having exposure to this transformational technology”. Even the most bearish house, BCA Research, remains neutral on equities precisely because of AI capex. Sixty voices, one direction.

There is something here worth examining more clinically, and it is this: if the outlook were an exercise in independent analysis, we would expect dispersion, not convergence. When sixty research teams, with access to different data, different models, different expertise, arrive at the same conclusion, there are two possibilities. Either they are looking at a truth so self-evident that everyone can see it, or they are all looking at the same point because looking elsewhere costs too much. The history of these documents suggests the second reading. The “investment outlook” function of the large private banks is not designed to challenge consensus but to legitimise it retrospectively; it is the mechanism through which the banker justifies to the client portfolio choices already aligned with the mainstream, lending them the form of strategic conviction. It needs to be said immediately that the people inside these institutions are not charlatans but competent professionals acting in good faith; the point is not the individual’s ill will, it is the architecture of incentives that converts individual good will into institutional convergence.

The pattern is verifiable by looking at the ESG case, because that one has now played out in full. From 2018 to 2024, global sustainable funds attracted steady inflows, the best year recording $38 billion net, with Europe becoming the epicentre holding 86% of assets. The outlooks of the major banks in those years were saturated with energy transition, transition investing, climate alpha, and whatever else with a green label could be assembled. Then in 2025 the signal inverted: $84 billion of global outflows, the first year of net redemptions since Morningstar began tracking the segment, 335 funds renamed in the first quarter alone to strip out words that had become inconvenient. Clients who entered at the peak are leaving at the trough, and the 2026 outlooks have already shifted vocabulary, now writing “selectivity”, “transition realism”, “multipolar optimism”. The same institutions that built the entry narrative are now building the exit narrative, and in between sits the actual damage done to the client.

This is not the first time I have watched the mechanism from the inside. In the early 2000s I worked as an adviser to a venture capital firm, analysing business plans, and during the months of the dot-com bubble they arrived in their thousands, under the explicit pressure of a pension fund that wanted a portion of its capital allocated to the tech theme. Most of those projects were threadbare: non-existent revenue models, teams without traction, valuations built on invented multiples. But something had to be picked, because the fund needed to show its contributors that it was positioned on the trend of the moment, and the VC needed to show the fund that it could deploy the capital. The pressure was not vertical from the top; it was a chain. Contributors asked “are we in internet?”, the pension fund passed the pressure down to the VC, the VC passed it to the analysts. Everyone, at their respective floors, was doing “analysis”; in fact, no one was doing analysis.

In the same period I sat on the board of a startup evaluating Magex, the NatWest spin-off built around digital payments and DRM for e-commerce content. Magex had raised £50 million from Goldman Sachs and Capital Z, with Universal Music and Reuters among its investors, and presented itself as probably the first serious payment gateway for British e-commerce, particularly for downloadable products such as music. The technology was sound; the problem was timing. The market for paid digital content was not yet there, the consumer infrastructure was not yet there, and Magex disappeared shortly after. NatWest had backed the project because in 1999 a British high street bank without a visible bet on digital commerce would have been accused of not understanding the internet. A wrong time-to-market was an acceptable cost relative to the reputational cost of being absent. The bank was responding to the same pressure as the pension fund, only towards a different audience: shareholders, the analysts covering it, and the financial press.

Eight years later, the dynamic repeated itself identically with subprime mortgages: products that the banks’ own internal risk officers were flagging as problematic were nevertheless distributed, because staying out of the mortgage securitisation market in 2006 meant losing share in a segment growing at thirty per cent a year. When the 2000 bubble burst and when 2008 arrived, no one was sacked for having been long the wrong themes; those who would have been sacked were the ones who had stayed out. The financial professional’s incentive is not to maximise the client’s return; it is to minimise the socially visible regret of his own institution.

When I was younger, I believed that banking prudence was a structural feature of the institution; I thought that the bank’s balance sheet, the prudential ratios, the risk management culture would form a barrier. I have learned that banking prudence is a feature of the individual banker when the individual banker can afford it, and that within the large private bank, few can afford it. Career risk is asymmetric: being outside consensus and wrong is fatal, being inside consensus and wrong is normal. It was true in 2000, it was true in 2007, it is true now. The portfolio of the client who follows every pop trend is not prudent; it is socially covered. The cover, however, is the banker’s, not the client’s.

This does not mean that AI as a theme is wrong; if anything, the capex of the seven hyperscalers is real, the $500 billion of annual investment announced for 2026 is real, the macro impact is real. It means something subtler and more operationally relevant. The client entering AI themes now is doing so in the third year of the cycle, after the consensus has formed, after valuations have already discounted collective conviction, after the same outlooks recommending exposure today are admitting that US equity market concentration is at all-time highs and that tech-plus represents nearly fifty per cent of the index. The problem is not AI as a thesis; it is the structural timing of entry, dictated by the fact that the outlook machine produces collective conviction precisely when the informational edge has already been eroded.

There is a further dimension that the UHNW client rarely registers, which is, for instance, that when ESG was the theme, the outlooks explained why it was impossible not to have exposure, and when ESG became problematic, the outlooks explained why it was reasonable to reduce exposure. In both phases the document was presented to the client as analysis; in both phases it was post-rationalisation of the bank’s socially sustainable positioning at the moment of writing. The client who read the 2022 outlook on transition themes and reads the 2026 outlook on AI themes today is reading the same object, produced by the same machine, with the same function: to make the bank’s presence visible, not its insight.

The falsifying evidence for this thesis would be the following: regularly finding, inside the quarterly outlooks of the major banks, recommendations that contradict the consensus of the moment, that turn out to be correct ex post, and that were maintained even when accounting to the client for a contrarian position was painful. Sixty years of documentation of the “investment research” function of the great international banks do not offer that evidence. They offer the opposite pattern: convergence to consensus, vocabulary recalibration when consensus changes, rare or no anticipatory protection of the client when consensus inverts.

Anyone who manages significant wealth and reads their private bank’s quarterly outlook should read it for what it is: a document of institutional positioning, not independent analysis. This does not mean that the research output of the major banks is useless; on the contrary, for ordinary portfolio management, execution, currency hedging, tactical allocation within defined ranges, the analytical infrastructure of the majors works very well. The problem appears only at the turning points of the cycle, which are precisely the moments at which the client would most need a non-aligned analysis. Independent analysis is expensive, because it requires accepting that the client may be positioned in a way that is not socially defensible for two or three years before the market proves it right. That service exists, but it is not the service that produces a graphically polished quarterly outlook distributed to ten thousand clients; it is structurally different. It operates outside the constraint of socially covered consensus, it explicitly accepts the reputational risk of being alone in being wrong, and for that reason it can afford to see the turns of the cycle before they become shared vocabulary. Those who have not built that channel, alongside the relationship with the private bank, will find themselves structurally exposed at the next turn, in exactly the same way they have been at the last three.